Ontology is the process of growth and elucidation of concepts of an information domain being common for a group of users. Establishing ontology into information retrieval is a normal method to develop searching effects of relevant information users require. Keywords matching process with historical or information domain is significant in recent calculations for assisting the best match for specific input queries. This research presents a better querying mechanism for information retrieval which integrates the ontology queries with keyword search. The ontology-based query is changed into a primary order to predicate logic uncertainty which is used for routing the query to the appropriate servers. Matching algorithms characterize warm area of researches in computer science and artificial intelligence. In text matching, it is more dependable to study semantics model and query for conditions of semantic matching. This research develops the semantic matching results between input queries and information in ontology field. The contributed algorithm is a hybrid method that is based on matching extracted instances from the queries and information field. The queries and information domain is focused on semantic matching, to discover the best match and to progress the executive process. In conclusion, the hybrid ontology in semantic web is sufficient to retrieve the documents when compared to standard ontology.
The difficulty of information storage space and retrieval has concerned escalating special treatment since 1940. The difficulty affirms that huge quantities of information to be stored and the relevant information should be precise. An enormous contract of research work has been completed to offer speedy and intellectual retrieval methods. To the research concern of digital libraries, several indeed contain information storage and retrieval troubles, such as logging and textual penetrating. Conversely, the difficulty of successful repossession continues mostly vague. Civilizing the usefulness is a significant ambition for the research of information retrieval system. Identifying the concepts or effort of the user is the major complicated obsession for relevant documents searching from a huge amount of information. For the user using common terms of queries for searching, an information retrieval system will not provide functional and detailed answers. The domain information of documents and cognition of the user are thus major for the retrieval of relevant documents information.
The research on combining the methods of ontology and information retrieval for semantic web is emerging in recent times. To explore the relevant information for the users need, a conventional method is introduced by entrenching ontology into information retrieval. If the investigated information is enclosed beneath the knowledge domain of user’s concepts, the motivation increases the probability of relevance. Therefore, the efficiency possibly enhanced. The challenges of implanting domain knowledge into information retrieval system are as follows. What is the apposite information retrieval model? How to execute and build ontology? How to discover the relevant documents by ontology?
The semantic web is build for current web extension where the information has well defined meaning and enabling cooperation between people and computers. Because of this well-defined structure, humans and even machines will work in coperation. The standard fuzzy ontology is a technique which is used in information retrieval where the calculation of relationship among the concepts are done using membership values. From domain’s uncertainty data, generation of fuzzy ontology automatically is highly desirable. This research explores hybrid fuzzy ontology-based information retrieval models in semantic web and gossip about the achievement and authority of applying proposed ontology containing common field knowledge and fuzzy concepts fabricated from the stored documents automatically. For mapping, the generated fuzzy ontology to semantic representation Web Ontology Language (OWL) is used.
This research work is organized as follows: The related work is reviewed in Section
Tho et al. [
Formica [
To design information retrieval system, the major challenges for researchers and developers is the method of sharing and searching the information with emergence of web. Kohli and Gupta [
Sometimes irrelevant information is retrieved on the semantic web but it is meaningful, and with ontology mapping, the relevance can be improved. Kandpal et al. [
Recently, the data originated from multiple types of sources includes the mobile devices, individual archives, sensors, social networks, enterprises, and cameras; Internet of things, software logs, and health data have led to one of the most challenging research concerns of the big data era. So, Xu et al. [
Liu et al. [
So, the study of related works motivates the semantic matching technique by combining the fuzzy ontology with keyword matching to retrieve the relevant information.
The hybrid ontology approach to query interpretation is on the aspiration of generating more than one specific planned query from a given keyword. This research refers to every produced query as an elucidation. The proposed model uses a hybrid fuzzy ontology for semantic relevant document retrieval. It semantically repossesses a position of related documents along with users query esteeming the emphasized sector or domain. It can be used to retrieve every category of documents in a particular domain written in all languages. The proposed information retrieval models and their major components are a set of annotated documents, user’s queries, retrieval engine, and ranking module. The relationships between concepts are built using ontology terms and NLP techniques. The relationships and natural-language synonyms represents the entities which completes the ontology by considering the key technique of NLP.
As demonstrated in Figure
Hybrid ontology for information retrieval.
Usually, for information retrieval system the documents are processed in two phases: document processing and query processing. In document processing stage, by using textual preprocessing the documents are processed to gain imperative stipulations and features for representing the documents. The conditions then are applied to construct fuzzy taxonomies from side to side of the ontology building techniques. The concepts contain definitions and instances which is given by the textual description of WordNet. WordNet can be satisfied as a moderately structured synonym store.
There are three databases in WordNet, noun is the initial one, verbs is second database, adjectives and adverbs are the final one. “Synsets” is a set of synonyms which designate a concept or a sagacity of a set of terms. Synsets available make diverse semantic relations for instance synonymy (similar) and antonymy (opposite), hypernymy (super concept)/hyponymy (subconcept) (also known as a hierarchy/taxonomy), meronymy (part-of), and holonymy (has-a). Depending on the grammatical category, the semantic relatives with the synsets will vary. The following sections discuss about document processing and information retrieval using standard fuzzy ontology framework.
In FOGA [
The traditional FOGA framework.
In Algorithm
Algorithm: Fuzzy Generation Input: Starting concept Output: A set of generated conceptual clusters Process: (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) endif (11) endfor (12)
To characterize vague information, the restriction of fuzzy logic will be integrated into ontology. Characteristically, fuzzy ontology is constructed from a predetermined concept hierarchy. On the other hand, a complicated and tedious process is assembling the concept hierarchy for a particular domain. To overcome this difficulty, the FOGA is implemented for generating fuzzy ontology automatically on information uncertainty.
Ducatel et al. [
The equal relationships with keywords is segregated for illustrating the semantic conceptions among documents in term comparison processing. The quantity of identical association can be calculated by semantic comparison calculating process. In a set of procedures, based on WordNet the word comparison between keywords is intended through similarity measure.
Initially in document investigation, the important keywords are selected from the documents as the specific keyword space
Algorithm: Keywords Matching Input: Keyword space Output: The presentation for document collections
As the keywords in
The similarity strings.
According to the professional field (computer science), the ontology model is constructed which is depicted in Algorithm
After splitting the query into meaningful words, each word should be checked against the ontology. The entire amalgamation of words is in use for processing. Scrupulous domain ontology is received to verify whether the declaration is to provide ontology. If persuaded, then the association of the words is obtained into the deliberation. The points are described for matching ontologies and the rules used to group related concepts together are listed below (parents-superset and child-subset). The parent conception demonstrates the perspective of the concept, from this parent each matching concept are collected. Matching concepts with similar parent are controlled by individual score, ought to be located jointly under individual score. Each series of parent-child associated matching concepts that demonstrates the context of the series must end in a non-matching concept. Unconnected groups are attached together as afforest, prepared by the highest score of the group. If the parents have the children with similarity then they will acquire the privileged of two portions and are connected together.
In consequence the amalgamation of mutual hybrid FOGA and keyword matching with an elucidation of a keyword query is set together by individually matching the query terms in the keyword query against the elements of annotation store. An annotation store
This research work envisages the following three forms of matches.
Algorithm
Algorithm for relevance path-match Input: match point = Output: bool = {PASS, FAIL}
if return(PASS) else if process & extract(match path) end if return(FAIL)
For instance, as the synonym “fone” is connected with the concept PhoneNumber, then TypePath index maps “fone” to the type PhoneNumber, to the path Author-Phone.phone, and so on. As such, the synonyms “callin,” “dial-in,” “concall,” and “conferencecall” are mapped to the type ConferenceCall. The keyword “tom” has a value match with Author.name, AuthorPhone.author.name indicating that “tom” has appeared as the name of the author of an email, as the name of a person who was declared in the signature block of an email, and so forth.
The next step of comparison measure retrieves and ranks the relevant documents from the document database. In the beginning, the ontology of query preferred form the initial step (in Algorithm
Input: Output: An ontology of query
If MR <
This section described the experimental setup for hybrid FOGA using keyword matching to retrieve the relevant information and ranking the documents automatically. The dataset is constructed using list of abstracts selected from 1000 documents which are all collected from the web. Initially the documents are updated to the FOGA framework with preprocessed information. The elimination of stop words and operations of stemming are performed. The weight estimation process is done with term analysis and semantic analysis tasks. The related journals are collected for the fuzzy ontology from the web. Using HTML, the abstract pages are intended for manuscripts. The text document conversion is done by removing the HTML tag elements from the web documents and document information is maintained in separate files. The two most common and important metrics for information retrieval efficiencies are precision and recall. In consequence, this research work used these measures for the ontology presentation for evaluation. Precision and recall are described in terms of a set of retrieved documents (e.g., the list of documents listed through a web search engine for an uncertainty) and a group of relevant documents (e.g., the list of every document on the net that is applicable for a convinced area):
The standard precision combines each query at recall level diagonally and calculates whole system performance approximately on a document/query capability.
For the sake of precision and recall, some researchers improve the architecture of inverted files. The authors move query keywords to semantic terms. But index tables still used keyword-based ones. To make the match easier, a new index table with semantic terms is proposed in this work.
The combination of standard ontology with FOGA techniques in this research, prescribes the solution for information retrieval using keyword matching indexing techniques. The
Both Figures
Showing the precision and recall for proposed hybrid FOGA.
The
To evaluate the proposed hybrid FOGA framework this research collected a set of 1,000 scientific documents in the research area “information retrieval.” There are two shortest goals general to all IR methods: (a) effectiveness: IR must be accurate (achieves what the user expects to observe in the answer); (b) efficiency: IR should be speedy (quicker than chronological scanning). The main goal of information retrieval is to possess relevant documents in response to user needs. The performance of ontology is evaluated with the research area hierarchy created using hybrid FOGA. Initially precision, recall and
The hybrid fuzzy ontology.
In this research, a latest approach for retrieving information successfully through implementation of hybrid ontology is discussed. This research presents a development in the hybrid ontology semantic information retrieval through (a) getting back a group of relevant documents semantic method using the proposed hybrid ontology, (b) dealing with the variety of field topics problem using hybrid concept view fuzzy ontology, and (c) ranking the end result set of documents according to
The authors declare that there is no conflict of interests regarding the publication of this paper.